We propose the development of a novel theoretical framework for the creation of next-generation quantum mechanical force fields (QMFFs) designed to meet the challenges of biomolecular and drug discovery applications. The proposed QMFFs are orders of magnitude faster than comparable linear-scaling electronic structure methods because the explicit interfragment orbital coupling is replaced by an empirical term that depends, at most, only on the electron density. The empirical parameters are tuned to reproduce experimental condensed phase properties such as ligand binding affinities, pKa shifts, solvation free energies, and lattice energies. In these ways, the proposed QMFFs will offer improved accuracy and predictive capability for multiscale modeling applications of complex biological and pharmaceutical problems. Specifically, we propose to 1. Create a general theoretical framework for development of new QMFFs for molecular simulations. This framework will allow arbitrary quantum mechanical Hamiltonian models to be integrated to form new QMFFs with accurate, balanced intermolecular interactions. Focus will be placed on the parameterization of intermolecular interactions using highly efficient low-level approximate density-functional tight-binding models, and high-level ab initio density-functional and coupled cluster methods that will be accessible through a specific interface with the PSI4 quantum chemistry package. The methods will take advantage of new linear-scaling methods for electrostatic interactions and generalized solvation effects, and will be implemented into a flexible set of software libraries that have interfaces with molecular simulation programs. 2. Apply QMFFs to thermodynamic properties of crystalline pharmaceuticals. The bioavailability of drugs are affected by crystal polymorph and co-crystal properties, such as solubility, and their thermodynamic characterization is thus an essential part of the drug discovery and chemical process pipeline. Density-functional QMFFs will be developed to study crystalline polymorphs by parameterizing them to reproduce known lattice energies. We will exploit the decomposition of the QMFF energy to compute the relative free energy between polymorphs, and absolute sublimation and solvation free energies of crystals using novel simulation methods that scale the intermolecular interactions. Our QMFF's ability to describe bonding events will be used to explore pKa shifts of drug compounds in co-crystals and make predictions about protonation states that are of pharmaceutical importance. 3. Apply QMFFs to aid in the interpretation of spectroscopic data. We will apply density-functional QMFFs to aid in the interpretation of IR, Raman, and NMR molecular spectroscopy of proteins and nucleic acids. First, small biologically relevant molecules will be used to benchmark the ability of density-functional QMFFs to predict IR, Raman and NMR spectra from molecular simulations. Second, we will apply the QMFFs to probe electric field heterogeneity on ketosteroid isomerase (KSI) with IR and 13C NMR chemical shifts. Third, we will examine the NMR chemical shift distributions in DNA and RNA systems derived from representative ensembles generated from MD simulation, and compare with recent and ongoing experiments.